A tale of two votes, Nigel Roberts

Towards Consensus? The 1993 General Election in New Zealand and the Transition to Proportional Representation
Jack Vowles, Peter Almer, Helena Catt, Jim Lamare and Raymond Miller
Auckland University Press, $39.95
ISBN 186940 123 9


“It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness … it was the season of Light, it was the season of Darkness….” The opening lines of Charles Dickens’ A Tale of Two Cities encapsulate not only the era of the French revolution, but also the mood in New Zealand in recent years.

Unbounded optimism and unbridled pessimism were the hallmarks of the proponents and opponents, respectively, of electoral reform during the 1993 referendum debate in New Zealand. And despite the fact that the 1993 election and electoral referendum were more than a year and a half ago, the confusion persists. A la Dickens, some people think we have everything before us; others argue we have nothing before us. Many are claiming that New Zealand made a drastic mistake on November 6 1993 when we adopted MMP, the mixed‑member proportional representation electoral system, and there have already been calls to relitigate the question by holding another referendum.

It is important to note, though, that the confusion and malaise that currently dog the body politic in New Zealand are not the result of MMP. They stem from two factors, neither of which can be blamed directly on proportional representation.

In the first place, the result of the 1993 election was exceptionally close. National won 35.1% of the valid votes and 50 of the 99 seats in Parliament. Labour was hard on its heels with 34.7% of the votes and 45 seats. The Alliance and New Zealand First won two seats apiece.

This was an even tighter result than the 1957 election, when Labour captured 41 seats compared with National’s 39; and closer than the 1981 election, when National squeaked back into power with a two‑scat majority. (On that occasion, the government won 47 seats in the House of Representatives, against Labour’s 43 and Social Credit’s 2.) The difficulties Sir Robert Muldoon encountered after the 1981 election should not be forgotten: his government failed to last a full three‑year term and National was defeated in the snap‑election that was held in July 1984.

As a consequence of its minuscule majority, it was thus always inevitable that Jim Bolger’s government would be under a great deal of pressure during the life of the 1993 to 1996 Parliament (assuming, that is, that the government manages to hang on until, and that an election isn’t called before 1996).

A second ‑ and possibly even more powerful ‑ cause of the disorder in the current political scene can be ascribed to the reduction in the number of electorates in the next Parliament. It should be noted, however, that this isn’t a consequence of proportional representation per se. “Downsizing” is an insidious piece of jargon, usually associated with corporate restructuring. During the past decade the process has caused a great deal of pain for tens of thousands of people throughout the country and the downsizing of the number of electorates from 99 to 65 was naturally going to do the same for current members of Parliament. To use the title of an old movie, Three Into Two Won’t Go, it was utterly predictable that the two larger parties in Parliament, National and Labour, would feel the squeeze.

For example, it was clear from the start of the redistribution process that Ross Meurant’s Hobson electorate was going to disappear (and it certainly did: Whangarei has eaten a goodly slice of it and Northland has swallowed the rest). So it was hardly surprising, then, that Meurant ‑ who found himself between a rock and a hard place ‑ should jump ship and form RoC, the Right of Centre Party. Likewise, more than half of Trevor Rogers’ electorate, Howick, has been incorporated in the redrawn Pakuranga and he, too, has walked the plank and ended up on the rocks.

On the other side of the House and at the other end of country, problems were unavoidable in Dunedin, which has long been a three‑seat citadel for the Labour Party. With three active, high‑profile MPs ‑ none of whom was ripe for retirement ‑ representing the city, the downsizing of Dunedin to two seats (now known simply as Dunedin North and Dunedin South) was always going to be a painful operation for Labour. This was dramatically proven to be the case in late June 1995, when Clive Matthewson, a former Labour cabinet minister, broke from the party he had represented in Parliament for 11 years to become leader of the United New Zealand group. He was accompanied by six other MPs ‑ four from National (Bruce Cliffe, Pauline Gardiner, Peter Hilt, and John Robertson), and two from Labour (Margaret Austin and Peter Dunne, the latter having come via Future New Zealand).

The extent of the disruptions to normal political viewing patterns in New Zealand can be underlined by examining what has happened to the government since November 1993.

A standard way of classifying governments is to ask whether or not they command more than half the votes in Parliament, and ‑ second ‑ whether the governments are composed of one or more parties. This leads to four basic types of government: single‑party majority governments; coalition majority governments; coalition minority governments; and single‑party minority governments. For many years New Zealand was listed as a classic example of a single-party majority government. For an example of a coalition majority government one need look no further than the current German government (or, indeed, any of its post‑war West German predecessors). Carl Bildt’s 1991‑1994 Conservative‑led coalition in Sweden is a clear example of minority coalition government (though composed of four parties, the government held only 170 of the 349 seats in the Swedish Riksdag); and Gro Harlem Brundtland’s long‑standing Labour administration in Norway is a good example of a single-party minority government.

In just over a year and a half New Zealand has experienced three of those four types of government. When it was sworn in after the 1993 general election National could command the support of 50 of the 99 seats in the House of Representatives and was thus a single-party majority government. In late 1994, when Meurant left National to form RoC, he retained his post as a parliamentary under‑secretary and thus as a member of the Executive. As a result New Zealand then had a majority coalition government (there was a formal agreement between the two parties which ensured that the government could rely on 49 National MPs and one RoC member for support on issues of confidence).

When Graeme Lee left the National Party to set up the Christian Democrats in May this year, the National‑RoC coalition lost its majority (it could command the support of only 49 MPs) and the New Zealand government thus entered its third phase since the last general election ‑ that is, it became a minority coalition government.

Bolger was loathe to admit this and was very irritable when journalists or political scientists had the temerity to suggest that he led a minority government. A month and a half later, however, he had no choice. The formation of the United New Zealand Party left the parliamentary line‑up as National 43 seats and RoC 2 seats (giving the government a total of 45 seats); Labour 42 seats (but this included Dr Peter Tapsell, who is the Speaker of the House); United New Zealand 7 seats; the Alliance 2 seats, New Zealand First 2 seats and the Christian Democrats 1 seat.

These changes to the types of government New Zealand has experienced since the 1993 election may well be causing confusion for spectators observing the political and parliamentary arenas, but we are probably fortunate they are occurring. It is a time of learning for everyone and a very good lesson to emerge from the seeming chaos is that majority coalition governments are not the only alternative to the standard Westminster single‑party majority government that New Zealand has grown so accustomed to during the past 60 years. In 1993 it sometimes seemed that Helen Clark was the only politician who understood that (and for good reason: both before she entered Parliament and since she has paid close attention to Scandinavian politics). Now even Bolger and Bill Birch have grasped the fact that governments can lose votes and not lose either sleep or ‑ more importantly ‑ office.

Despite the turmoil currently afflicting the political scene in New Zealand, however, one thing that all the parties contesting the 1993 general election and both sides in referendum debate would probably agree on is that the 1993 election and referendum were a watershed in the politics of this country. They changed the political landscape ‑ probably forever ‑ and Vowles, Aimer, Catt, Lamare and Miller are to be congratulated on producing such a detailed and timely study. The authors and their publisher are also to be commended for the speed with which Towards Consensus? has been produced. We had to wait almost three years ‑ nearly the full length of the first term of the fourth National government ‑ for Voters’ Vengeance (Vowles’ and Aimer’s study of the 1990 election) to appear; its sequel, by way of contrast, is available a mere 20 months after the 1993 election and referendum.

Towards Consensus? is a book that everyone who has more than a superficial interest in New Zealand politics will have to buy. It will be compulsory reading for all political activists and analysts. Given the widespread interest overseas in New Zealand’s electoral experiment, the book should deservedly be assured of a place on the shelves not only of public and university libraries, but also of politicians and political scientists.

Jack Vowles and his collaborators tell us not simply what happened in the 1993 election and referendum but, more importantly, why. To do so, they draw on data from a post-election mail questionnaire sent to just over 3000 registered electors, with a very creditable response rate of 73%. In addition, they also sent to all four main parties’ parliamentary candidates and national conference delegates a 26‑page questionnaire that contained many of the questions they put to their sample of registered electors. Somewhat surprisingly, the response rate for this “elite” survey was only 60% ‑ notably lower than that for the “mass” survey. The value of the mass survey was further enhanced by the fact the very nearly half the respondents were people to whom Vowles and Aimer had sent questionnaires three years previously as a part of their 1990 postal survey.

As the pre‑eminent British statistician, Sir Claus Moser, has noted, “mail questionnaires addressed to the general population tend to result in an upward bias in social class composition and educational level”. As a result, it’s not surprising to find that Maori are under‑represented in the mass survey: for example, they account for less than 7% of respondents in the analysis by Vowles et al of the way in which different social groups voted in the referendum. Maori also account for less than 4% of the elites in the same table and there are “too few cases” of Pacific Island elites to be included in the analysis. These statistics are an inevitable consequence of the fact that New Zealand is a divided, stratified society, but ‑ by and large ‑ Vowles and his colleagues are able to counter the worst effects of under-representation by weighting their data when necessary to ensure comparability with known facets of both the 1993 election and referendum results.

Significantly, Towards Consensus? documents the fact that “the historical dominance of National and Labour was waning even before its institutional foundations in the first-past‑the‑post electoral system were further weakened by the referendum in 1993”. In other words, what happened in 1993 may reasonably be viewed not so much as a single decisive break with the past, but ‑ rather ‑ as yet another milestone along a constantly changing road.

Of course, some things change more quickly than others and Vowles, Aimer, Catt, Lamare and Miller rightly pay attention not only to the “ties that bind”, but also to the causes of “volatility”. On the one hand they find that in 1993 New Zealand’s party system was “still essentially based on the traditional but weakening social‑structural differences which have separated support for Labour and National since the 1920s and 1930s…. [E]mployment continues to structure electorate outcomes and this effect actually strengthened slightly in 1993. The other traditional cleavage, that of class or occupational status, has weakened over the last 30 years but in 1993 it showed signs of a weak recovery after a deep trough in 1987 and 1990.”

On the other hand, however, the authors of Towards Consensus? find that “at the most only one‑third of the present electorate has a history of voting for the same parties in three consecutive elections”. The pattern of voting behaviour in New Zealand, they suggest, resembles a dance, but one in which the “electoral choreography is loose and open to much improvisation. Where voters get to on the electoral dance floor by the time the music stops on election night and how they get there is very much a matter of individual choice.”

In view of the fact that “the process of radical economic reform” initiated by the fourth Labour government “fundamentally redefined the relationship between … voters and their representatives” and the fact that the National government that replaced the Lange‑Palmer‑Moore Labour government had a legacy of “policy U‑turns, broken promises and outright deception” it is hardly surprising that the dance steps of voters in New Zealand have owed more to St Vitus than Strauss. Indeed, Vowles, Aimer, Catt, Lamare and Miller go so far as to argue that “we can expect a high rate of volatility to persist at the first MMP election when the presence of more parties will offer hitherto unmarked pathways of electoral escape and dispersal for a fickle electorate”.

This conclusion on the part of the authors of Towards Consensus? in turn points towards what is probably the most interesting chapter in their book ‑ that is, to their final chapter, which is called “Towards MMP”. More than any other section, this will be required reading for amateur and professional political analysts alike. An American political scientist who knows New Zealand well, Professor Jack Nagel of the University of Pennsylvania, has examined the international relationship between electoral systems and party systems and has suggested that under MMP New Zealand is likely to have between five and six parties. Incidentally, this should serve as something of a warning to the seven parties who are represented in Parliament at the moment (that is, in the last week of July; there could be more by the time this review is published), let alone to the hopefuls who have registered with the Electoral Commission but are not yet either in Parliament or even above 5% in the opinion polls.

Vowles et al use cluster analysis to create four, five, six and then seven different groups of electors and they locate them in a two‑dimensional space based on an economic left/ right axis and a social liberal/ conservative scale. If, for example, five ideological clusters are created one finds that there is a centre‑left constituency and a centre constituency, each containing about a quarter of the electorate, a liberal‑left group that’s almost as strong (22%), and liberal‑right and right groups of 17% and 11% respectively. However, it’s only when the authors of Towards Consensus? massage their data to achieve seven ideological clusters among New Zealand voters that a liberal‑centre constituency ‑ along the lines of the new United group in Parliament ‑ appears, and even then it contains only 6% of the electorate, just enough to cross MMP’s crucial hurdle.

Vowles and his colleagues candidly admit that “the implications of this information for tomorrow’s parties is [sic] by no means clear”, but they are prepared to offer some sensible speculation. They say that “if ACT took all 1993 Labour votes from the liberal‑centre and right clusters it might just surmount the 5% threshold. Given probable support from former National voters, this is the party most likely to succeed in making the next election a five‑horse race. ”

Strategic voting like this is a phenomenon of proportional representation systems. For example, it is a well‑documented fact of Swedish political life that there are Social Democrats who have deliberately voted for the Left Communist Party (a group which rapidly renamed itself simply the Left Party after the fall of the Berlin Wall!) both to help the party clear Sweden’s 4% hurdle and thus ensure the Left’s representation in Parliament, and also ‑ to use Don Chipp’s famous phrase from Australian politics ‑ “to keep the bastards honest”.

When all is said and done, that’s what the 1993 general election and the binding electoral referendum which accompanied it were about.

Nigel Roberts has taught political science in New Zealand universities and surveyed the views of voters in New Zealand general elections for more than a quarter of a century.

Tagged with: , , , , , , , ,
Posted in Non-fiction, Politics & Law and Review
Subscribe to NZ Books
We're pleased you're using the New Zealand Books archive.

To ensure the survival of this important journal, please consider
subscribing — only $44 a year, or $30 for digital-only.

Go to the Subscribe page.
Search by category

Read more